Before he died, Gordon Clark published an article The Trinity Review in which he summarized his objections to empiricism.
Let’s examine his objections:
"The theory of knowledge that presumably accords
best with common sense is the theory that we learn
by experience. We learn that bees sting and
rattlesnakes kill through our perceptions of pain.
We learn that roses are red and violets are blue by
the sensations of sight. All our knowledge comes
That’s a straw man argument insofar as one doesn’t have to believe the senses are the source of all knowledge to believe the senses are the source of some knowledge. While empiricism entails sense-knowledge, sense-knowledge need not entail empiricism.
This distinction is important to keep in mind since Clarkians constantly conflate sense-knowledge with tabula rasa empiricism, then imagine that by disproving tabula rasa empiricism they thereby disprove sense-knowledge.
But it’s quite possible for a thinker to believe that some knowledge is innate while the senses mediate other kinds of knowledge. Indeed, these can be complementary positions.
For example, we can enumerate concrete objects because we enjoy an innate knowledge of math. At the same time, we don’t enjoy an innate knowledge of the concrete objects we enumerate. So you need both types of knowledge to complete the operation.
“However plausible this theory may be, it raises
some exceedingly difficult questions. For the
moment let us set aside the complexities in trying to
rise from fleeting sensations to the knowledge of the
incorporeal and eternal God. Instead, let us first
attend to the most simple parts of empiricism.”
He doesn’t bother to explain why that’s a problem. However, one of the standard theistic proofs treats what is fleeting as evidence for the need of something that isn’t fleeting to ground what is fleeting. So that’s not a problem belief in God. To the contrary, that’s an argument for God. What is fleeting is contingent. What is contingent is ultimately contingent on something that is necessary.
“Let us start with the red of a rose and the blue of a
violet. First, a description of sensation will show
that it does not give knowledge so readily as
common sense imagines. Not everybody sees roses
as red and violets as blue. There are some people
who we say are color blind, and there are degrees of
color blindness. It is difficult to tell what is color
blindness and what are color illusions. The real
color is very hard to settle upon. The condition of
the organ, the eye, a disease, temporary sickness, a
headache or extreme sensitivity change our color
How would a genetic defect like color blindness prevent color from being an object of knowledge? If you’re color-blind, the color of an object is not an object of knowledge for you, but that doesn’t mean it can’t be an object of knowledge for someone with normal vision.
Why would we delimit the scope of knowledge by reference to a genetic defect? By that token, would we also say the past can’t be an object of knowledge because some people suffer from senile dementia?
“Let me give you one little example. If you would
take a course in art, oil painting, you might take a
square of canvas and put some color paint on the
top half of it and another color on the bottom. It
could be red and blue or any two colors you wish
just so long as they’re different. And then after they
have dried, take a brush full of gray paint and just
bring it down vertically over the two parts of the
square and you will see that that one stroke of brush
has put two different colors on the canvas, the color
of the gray at the top is not the color of the gray at
the bottom half of the canvas. So the color that you
see depends on the background against which you
see it. And since there is always a background, you
never see anything as it is all by itself.”
Why does knowledge of color depend on seeing a color as it is “all by itself”? In the nature of the of the case, color perception involves the appearance of an object.
“I could also mention some optical illusions: the
Texas rancher who was sure he was seeing a mirage
and drove his pick-up truck into a lake.”
How could Clark identify an optical illusion in the first place if the senses are totally unreliable? To detect the difference between a mirage and a lake assumes the basic reliability of sensory perception.
“Some of my friendly opponents try to meet my argument against empiricism by claiming that I merely parrot the
ancient skeptics. I’m afraid of two things: The ancient skeptics didn’t know anything about Texas, and, in the second place, if I am parroting the ancient skeptics, that is not a sufficient answer to their arguments.”
i) True, but I just dealt with his example.
ii) To the extent that a Scripturalist is raising objections to sense-knowledge, he needs to limit himself to Scriptural objections. For a Scripturalist to raise extrascriptural objections to sense-knowledge would be self-defeating.
“Take one thing that certainly the ancients didn’t
know. Get a nice piece of bristle-board cardboard
and paint one-half of it with black India ink. Leave
the other half white and then put little swiggles of
black on the white half. Then get something that
will rotate at about 500 revolutions a minute, and
what color will you see? Will you see black? Will
you see gray? Well, if you haven’t done this
experiment I’m pretty sure you just don’t know. I’ll
tell you: You’ll see purple; you’ll see red; you’ll see
green; you’ll see some sort of brown. You will see
all these colors just from a mixture of black and
white, and this gives you considerable difficulty in
trying to say that you see the color of anything at all
or to paraphrase a little bit from Augustine, there is
nothing given (das Gegebenes, if you know the
German technical term), nothing given in sensation
without intellectual interpretation.”
i) Once again, color perception concerns the appearance of objects. Naturally the appearance of an object varies with the empirical conditions under which it’s viewed (e.g. background, lightening). That, however, doesn’t mean the color of an object can’t be known. Rather, an object will appear one way under certain conditions, and another way under different conditions.
For example, the apparent scale of a mountain is observer-relative. Mountains look smaller at a distance. Would we conclude from this fact that mountains have no objective size?
ii) The appearance of the object is still a determinate object of knowledge. The appearance varies with the determinants, but that doesn’t mean it can’t be known. While color perception has a subjective dimension, it also has an objective dimension. For instance, the composition of the surface will reflect or absorb certain wavelengths of the visible spectrum.
iii) Even at a subjective level, human beings are designed to register certain color qualia. Barring genetic defects (e.g. color-blindness) we share common qualia.
“And just to protect myself from these people who
think I’m as old as the Greek skeptics – I am getting
a little ancient, but I’m not quite 2,000 years old, I
guess I’m about 95 or something like that – but I
was traveling along the road from St. Louis to
Indianapolis on one occasion. This was before the
interstate was there, and as I looked ahead, I saw a
small truck standing by a barn. This was
approximately 1,500 or 2,000 feet ahead of me. And
it wasn’t a passenger car, it was a truck because the
front and the back were both vertical. There was the
truck standing by the barn. Now as we drove along
– and going at 75 m.p.h. you cover a few feet pretty
quickly – this truck suddenly became a mailbox on
a post. Now was it a truck or was it a mailbox?
Well, that depends on how far away from it you are.
And time forbids the multiplication of such
examples. Suffice it to say that they soon become
overwhelming. You have trouble with sensation.”
i) Now Clark is equivocating. He fails to distinguish between sensation and perception. He didn’t see a truck. He saw an ambiguous object–ambiguous due to the lack of visual resolution at that distance. He then interpreted what he saw as a truck. Perception involves an element of expectation.
ii) Moreover, Clark’s illustration is self-defeating since it ultimately depends on his ability to distinguish between what the object appeared to be at a distance, and what the object really was when he saw it close up. The comparison only works if he can distinguish between appearance and reality.
iii) Furthermore, as rational agents, human sensory perception was never meant to operate in isolation to human reason. Reason can correct for misperceptions. For example, M.C. Escher was fond of depicting optical illusions. We know these depictions are physically impossible. That’s part of their charm.
iv) Incidentally, isn’t it unethical for a man who denies sense-knowledge to get behind the wheel of a car? Given his denial of sense-knowledge, does Clark think there’s no difference between driving sober and driving drunk? Does Clark think it makes no difference which side of the road you drive on?
“In the second place, this empirical theory, after
making such a poor beginning with sensation,
requires a theory of images to account for the
retention of knowledge after the sensation has
stopped. When you talk about the sensation, when it
is gone, and you have an image that is retained,
there are other difficulties.”
i) Why that’s a difficulty, Clark doesn’t say. Does he mean reliance on memory is a difficulty, or reliance on remembered mental images in particular?
a) If the former, the rationalist has to rely on his own memory as well. For example, logical inference requires a recollection of the premise.
b) If the latter, what’s the problem? As long as my memory of what I saw (or heard or felt or tasted or smelled) corresponds to what I saw at the time, is there some additional problem?
ii) Why does Clark think empiricism or sense-knowledge depends on mental imagery, per se? It’s possible to remember that you saw something even if you don’t remember what you saw (in the sense of retaining a mental image of the event).
iii) Why does Clark restrict recollective mental representations to imagery? We can also have mental representations of auditory, tactile, olfactory, and gustatory sensations.
“If perception is an inference from sensation, and images follow the perception, how can one determine when the
inference is valid?”
This question is so wide-open that it’s hard to tell where he thinks the problem lies. I spot a snakeskin in my backyard. From this I infer the existence of a skin-shedding snake. What’s wrong with that inference?
“At one time, I inferred that I saw a truck. Another
time, a few minutes later, I inferred that I saw a
mailbox. But how do you tell whether either
inference is valid?”
But Clark himself tells us which is which. And he gives the reason. In this case, relative distance between the observer and the object was the differential factor. It also depends on visual acuity.
“And then in the second place, some people, especially scientists, not artists, but especially scientists, don’t have any images. And that’s a difficulty I don’t see how the empirical philosophy can ever overcome. They seem never to have thought of the existence of such people. Thomas Aquinas and David Hume, best known for their theories of images, just seem to believe that all people have images. But that isn’t so. There are some people, and I know one fairly well, who have no images at all.”
i) How is the absence of a particular sensation a difficulty that empiricism is supposed to overcome? Some men are blind. Some men are deaf. How does the inability of some individual to see or hear because they lack functional sense organs amount to an argument against the possibility of sense-knowledge for percipients who do enjoy normal sight or hearing?
Some males suffer from a birth defect known as aphallia or penile agenesis. Does this condition invalidate the science of andrology?
ii) How does deficient sensation in one sensory organ invalidate sense-knowledge generally? If you’re blind or color-blind, that doesn’t mean you can’t hear or touch or smell. And, as we know, different animals vary in their sensory aptitudes and acuities.
“Now, third, even for people who have visual or
auditory images, the formation of concepts by
abstraction, as Aristotle and Locke require, is
impossible for reasons I won’t go into. And if
Bishop Berkeley did nothing else, at least he clearly
showed that empiricism cannot allow or justify
Since he won’t go into it, there’s nothing to refute.
i) Why can’t empiricism justify abstract concepts? Does he mean empiricism can’t justify abstract universals? But even if enumerative induction can’t justify a universal inference, that doesn’t mean you can’t form an abstract concept on the basis of enumerative induction. There’s a difference between an abstract concept and an abstract universal.
If I hear a clock chime four times in a row, I form the abstract concept that it’s four o’clock. Several elements go into the formation of that concept: memory; auditory perception; a concept of numerical relations.
Now, I happen to think that our knowledge of numerical relations is innate. I don’t think empiricism can account for mathematical knowledge.
But, as I’ve said before, belief in sense-knowledge doesn’t commit you to empiricism.
“My fourth objection to empiricism, and if you’ve
been counting them up, it may be the fortieth,
empiricism cannot produce norms of any kind. It
cannot produce moral and religious norms because
at the very best, empiricism can only tell you what
is. I don’t think it tells you even that little, but that
is all that empiricists can legitimately claim to do.
They cannot tell you what ought to be because you
cannot get an ought out of an is. And this applies
not only to moral and religious norms, but to the
very basic logical norms without which speech and
understanding would be impossible.”
That may well be true, but how is that an objection to sense-knowledge? Even if moral norms are either innate, revealed, or both, we must apply moral norms to sensory objects.
Consider the Biblical prohibition against theft. Suppose I own a red Alpha Romeo. Suppose my neighbor owns a yellow Alpha Romeo.
Differentiating the cars by color is one way of differentiating my property from his property. Suppose we both drive to the store. I park my car next to his.
Would Gordon Clark take the position that when I return from the store, I should pay no attention to the color of the car? If my neighbor left his keys in the car, does it make no moral difference whether I drive away with the yellow Alpha Romeo or the red Alpha Romeo?
Or do I have a moral obligation to distinguish my car from his on the basis of visual cues–like color?
“The argument is that every philosophy must have a
first principle, a first principle laid down
dogmatically. Empiricism itself requires a first non-
empirical principle. This is particularly obvious in
that most extreme form of empiricism called logical
positivism. To say that statements are nonsense
unless verifiable by sensation, is itself a statement
that cannot be verified by sensation. Observation
can never prove the reliability of observation. Since,
therefore, every philosophy must have its first
indemonstrable axiom, the secularists cannot deny
the right of Christianity to choose its own axiom.”
But if a Clarkian says you can’t know something unless you can account for something (i.e. give reasons), then an indemonstrable axiom can’t be known to be true.
If, therefore, we treat the Christian faith as an axiomatic system, we can’t know that it’s true. And there’s no reason to prefer it to a rival axiomatic system.
“The principle is sola Scriptura. This is a repudiation
of the notion that theology has several sources such
as the Bible, tradition, philosophy, science, religion,
or psychology. There is but one source, the
Scriptures. This is where truth is to be found. Under
the word truth there is included, in opposition to
irrationalism, logic and the law of contradiction.
Whatever contradicts itself is not truth. Truth must
be consistent, and it is clear that Scripture does not
both affirm and deny an atonement. God is truth.
Christ is the wisdom and Logos of God. And the
words he has spoken to us are spirit and are life.”
i) Given his denial of sense-knowledge, Clark needs to explain how the Bible can even be an object of knowledge.
ii) I don’t deny that truth excludes contradiction. However, notice that Clark made no attempt to exegete the law of contradiction from the Biblical concept of truth. You can’t simply read that off the occurrence of the word “truth” in Biblical usage. What Clark has really done is to begin with preconceived notion of what truth entails, then map that back onto the occurrence of the word in Scripture. He didn’t get that from Scripture alone.
iii) He also has a bad habit of equating the Greek word “logos” with the concept of “logic.” But that’s not what the word “logos” means in Johannine usage.
“The first passage for exegesis is the first passage in
the Bible. God created man after His image and
likeness. This image cannot be man’s body for two
reasons: First, God is spirit and has no body;
second, animals have bodies but they were not
created in God’s image. Therefore, the body cannot
be the image of God.”
That’s a fallacious inference:
i) An image is a physical representation. The fact that what the image represents may be incorporeal doesn’t mean the image itself is incorporeal.
ii) In context, the imago Dei is a finite manifestation of God’s dominion. Man is God’s representative or vice-regent on earth. The exercise of dominion is something he shares in common with his Maker.
iii) Apropos (i)-(ii), the human body is an instrument of dominion. Man can exercise dominion over the sensible world because he has a body. So his body is a necessary expression of that principle.
iv) This is not to deny that we also need certain mental properties to execute the dominion mandate. But Clark’s interpretation is too narrow to do justice to the passage.
“The divine image then must be man’s spirit, for the two elements which compose man are body and spirit. Genesis says that God formed man of the dust of the ground and breathed into his nostrils the breath of life, and out of these two elements man became a living soul. If the dust or clay is not God’s image, the breath or spirit must be. There is no other possibility.”
Once again, that’s faulty exegesis:
i) In Hebrew usage, man doesn’t have a “soul”; rather, man is a “soul.” A “living being” is not a synonym for “spirit.”
ii) The “breath of life” simply distinguishes a corpse from a living organism.
I myself subscribe to dualism. But you can’t get dualism from these passages. For that you must look elsewhere.